Reviewed by Lexie CornerMay 7 2024
Scientists from the Genome Institute of Singapore (GIS) and Bioinformatics Institute (BII) at A*STAR have created a new artificial intelligence (AI) software tool named "BANKSY" that can recognize different cell types found in tissues, including fat, muscle, and nerve cells. This research was published in the journal Nature.
Beyond traditional AI technologies, which cluster cells together based on shared molecular structures, BANKSY considers the similarity of the cells' environment within the tissue. With BANKSY, researchers could more quickly and correctly gain a better understanding of tissue processes in various diseases, which could aid in creating more potent cancer, neurological, and other disease-related diagnoses and treatments.
BANKSY recognizes slightly differentiable cell groupings in spatial molecular profiles produced from tissue samples. It also addresses the unique but related challenge of defining functionally distinct anatomical regions in tissue sections. For example, it can discern between layered structures in the human forebrain.
Spatial molecular profiling, also known as Spatial Omics, encompasses powerful microscopy technologies that enable scientists to scrutinize tissues with remarkable detail. These tools unveil the precise locations of individual biological molecules within cells and the spatial organization of cells within tissues.
This aids in their comprehension of how cells interact in tissues to carry out regular physiological processes. It also helps them understand how cells behave or misbehave in conditions like cancer, autism, and infectious diseases like COVID-19. This knowledge is critical to the development of novel medications as well as more precise diagnosis and patient-specific treatment.
With the latest developments in Spatial Omics technologies over the past few years, BANKSY can assist biologists in interpreting and deriving insights from these technologies. Compared to other techniques, BANKSY is more adaptable, precise, quick, and scalable when analyzing RNA and protein-based spatial genomics data.
Compared to rival approaches tested, BANKSY is two to 60 times more scalable and can handle big datasets of over two million cells. It is also 10 to 1,000 times faster. This implies that the technique can be used for other important phases in data processing, like identifying and eliminating low-quality regions from the sample and combining samples obtained from several patients for a single study.
Two separate studies have independently benchmarked and determined that BANKSY is the best-performing algorithm for spatial omics data; one of these studies indicated that BANKSY can be a potent tool for domain identification. After putting six algorithms to the test, the other study determined that BANKSY was the most accurate for their data analysis.
We anticipate that BANKSY will be a game-changing tool that helps to unlock the potential of emerging Spatial Omics technologies. This will hopefully improve our understanding of tissue processes in diverse diseases, allowing us to develop more effective treatments for cancers, neurological disorders, and many other pathologies.
Dr. Shyam Prabhakar, Senior Group Leader, Laboratory of Systems Biology and Data Analytics
Dr. Shyam Prabhakar is also the Associate Director of Spatial and Single Cell Systems at A*STAR’s GIS.
Liu Jian Jun, Professor and Acting Executive Director at A*STAR’s GIS, said, “The work on BANKSY advances our strategy of combining high-throughput technologies with scalable, robust AI software for problem-solving and identifying the clues to what can make a difference in the lives of patients.”
We are using BANKSY to identify the cells that help tumors grow and spread to other parts of the body–drugs targeting such cells could be a promising direction for cancer treatment.
Dr. Iain Tan, Senior Consultant, Division of Medical Oncology at National Cancer Centre Singapore
Tan is also a Senior Clinician Scientist at A*STAR’s GIS Laboratory of Applied Cancer Genomics.
Journal Reference:
Singhal, V., et al. (2024) BANKSY unifies cell typing and tissue domain segmentation for scalable spatial omics data analysis. Nature Genetics. doi.org/10.1038/s41588-024-01664-3.